Remanenzflussbestimmung für das kontrollierte Einschalten von Leistungstransformatoren

In the last years controlled switching became an effective method to reduce switching transients which is documented by several publications – especially of CIGRE. Controlled switching is most commonly used for switching capacitor banks. Even though uncontrolled energisation of power transformers can lead to very high inrush current transients whose maximum values can exceed many times the rated current and cause enormous current forces in the transformer windings, controlled switching of transformers is rarely used in the field. This is mainly due to the fact, that additional voltage transformers have to be installed close to the power transformer in most cases, so that the residual fluxes – the optimal closing times are calculated based on these values – can be evaluated. On the one hand these sensors increase the costs of controlled switching and on the other hand the uprating of existing substations with controlled transformer switching is almost impossible. Thus, new methods to determine the residual fluxes based on the existing sensor system are presented in this work. First the required accuracy of the device to determine the residual flux has to be evaluated. It is shown that for three-legged transformers – the most common configuration of power transformers on the transmission level – only the behaviour of the first phase to be energised has to be analysed if the “Delayed Closing” strategy is used. An implementation of the “Rapid Closing” strategy in the field is hardly possible because the behaviour of the dynamic fluxes – they must be used to calculate the optimal closing time of the second and third phase to be energised – is unknown in most cases. Using the theoretical model of a single-phase transformer or simulations of a three-legged transformer, the required accuracy for the determination of the residual flux can be calculated. The results show that even with an ideal residual flux measurement the energisation of a transformer without inrush-currents in the field is not always possible. If a certain inrush current can be tolerated (here 1.0 pu), much more realistic requirements for the determination of the residual flux can be obtained that depend on the reserve of the transformer core utilisation as well as on the maximum closing time deviation of the circuit breaker. In worst-case an accuracy of 9 % is necessary. The residual fluxes are built up during the de-energisation process which normally occurs during steady-state no-load operation for the power transformers on the transmission level. At this time only the magnetising currents flow that amount to some amperes in maximum. Modern circuit breakers will chop these currents immediately after mechanical contact separation as well, whereby a transient ringdown phenomenon occurs on the load-side of the circuit breaker (current chopping). Due to these transients the magnetic fluxes still change after current interruption in the circuit breaker and thus the residual fluxes are influenced significantly. The three parameters that mainly affect this process are the instant of current chopping, the magnetising inductance of the transformer core and the capacitances of the elements located between the circuit breaker and the transformer (substation capacitance). As could be shown with a sensitivity analysis of a representative substation configuration, a generally valid behaviour of the residual fluxes depending on the instant of current chopping cannot be declared. Hence each substation has to be investigated separately. The simulation model used for these analyses can be adopted from the energisation studies and has to be completed with the capacitive elements solely. Additionally a scaling factor has to be taken into account for the simulation results because the modelling of the magnetisation behaviour is not optimal. A verification of the results with a 400 kVA transformer in the laboratory shows a good correlation between measurement and simulation for the energisation as well as for the de-energisation studies.